Carol R. Townsend

Natural Hybridization Between the Teiid Lizards Cnemidophorus tesselatus (Parthenogenetic) and C. tigris marmoratus (Bisexual): Assessment of Evolutionary Alternatives

Abstract Annual hybridization is taking place between representatives of the parthenogenetic lizard Cnemidophorus tesselatus (2n = 46, 47) and males of the bisexual species C. tigris marmoratus (2n = 46) in desert grassland habitats at Arroyo del Macho, Chaves County, New Mexico. This raises the question of whether a new triploid parthenogenetic species may be originating as a consequence of this activity. Hybrids were collected in each of four years (1996–1999), and 20 of 21 hybrids collected (12 males and 8 females) were available for study. Although a triploid parthenogenetic species (Cnemidophorus exsanguis 3n = 69) and a diploid bisexual species (C. inornatus 2n = 46) were also found at the hybridization site, the genealogy of the hybrids was determined unequivocally with karyotypic and electrophoretic evidence (34 loci tested). The specimens examined electrophoretically included an adult female and one of her laboratory-reared daughters, which demonstrated for the first time clonal inheritance in C. tesselatus pattern class E. The population of C. tesselatus at Arroyo del Macho is characterized by two karyotypic cytotypes. The ancestral one (2n = 46) occurs at about half the frequency of the derived cytotype (2n = 47), which apparently was produced by centric fission of the ancestral X-chromosome from C. tigris In contrast, the occurrence of the two cytotypes was reversed and strongly asymmetrical in the hybrids; only one of nine hybrids possessed the fissioned X-chromosome. This individual was significantly different in 12 meristic characters from the sample of hybrids with intact X-chromosomes. Predictably, principal components scores for this individual fell outside the 95% confidence ellipse of scores of the other eight hybrids that were karyotyped. The skewed ratio and multiple phenotypic differences suggest that hybrids inheriting a fissioned X-chromosome might be at a selective disadvantage compared to hybrids with intact X-chromosomes. All 20 hybrids closely resemble C. tesselatus in most color pattern features. However, these hybrids, like C. tigris marmoratus lack lateral stripes. Because the population of C. tesselatus at Arroyo del Macho has lateral stripes (or their remnants), hybrids can be readily distinguished from C. tesselatus by this color pattern feature. Compared to the two parental species, hybrids had a significantly lower mean number of scales around midbody, but hybrids resembled either C. tesselatus or C. tigris marmoratus in other univariate meristic characters. This mosaic pattern of resemblance was simplified to a three-dimensional depiction of variation using principal components analysis. Each of two principal components expressed the resemblance of hybrids to one of the two parental species. A third component reflected the difference between hybrids and both parental species. A canonical variate analysis of meristic characters demonstrated the multivariate distinctiveness of each group—hybrids, C. tesselatus and C. tigris marmoratus However, based on Mahalanobis D2 distances, the closest morphological resemblance among hybrids and parental species was between hybrids and the maternal species, C. tesselatus Nine additional museum specimens, suspected of being C. tesselatus × C. tigris marmoratus hybrids, were identified, as such, by a canonical variate analysis using our samples of C. tesselatus, C. tigris marmoratus and hybrids from Arroyo del Macho as a priori groups. These nine individuals document hybridizations between C. tesselatus and C. tigris marmoratus at two additional localities in Chaves County, New Mexico, two localities in Sierra County, New Mexico, and a cluster of sites near Presidio, Presidio County, Texas. Previously, several of these hybrids had been misidentified as male C. tesselatus The reproductive systems of female and male hybrids were compared histologically to those of C. tesselatus and C. tigris marmoratus respectively. Sexually mature and reproductive adults of C. tesselatus usually have oocytes in the ovary, complete and well-organized ovarian follicle walls, inconspicuous connective tissue and fewer vacuoles in the well-vascularized ovary, the distal oviduct with a thin mucosa, well-developed alveolar glands restricted to the middle oviduct, a proximal oviduct with a thick mucosa and well-developed folds, and small mesonephric tubules. Female hybrids have a poorly defined follicular epithelium with little vascularization in small ovaries, empty or fluid-filled follicles without oocytes, few or no cilia in the middle oviduct, and numerous abnormally large mesonephric tubules. There is no evidence that Cnemidophorus tesselatus × C. tigris marmoratus females can produce viable and fertile eggs. Although hybrid males are capable of producing sperm that appear normal and were present in the epididymides, the allotriploid chromosome complement reduces the chance that sperm would carry genetically balanced sets of information. Although the annual production of hybrids could affect the long-term success of this local population of C. tesselatus two lines of evidence indicate that hybridization is unlikely to result in its extirpation. First, the population of C. tigris marmoratus at Arroyo del Macho is tightly associated with a microhabitat dominated by creosote bush. Because creosote bush is distributed there in small, widely scattered patches, the density of C. tigris marmoratus is relatively low, and many individuals of C. tesselatus escape insemination. This was evident from an absence of sperm in the reproductive tracts of 11 individuals of C. tesselatus collected during the peak reproductive season (May and June) of three different years. Second, reproductively mature individuals of C. tesselatus are significantly larger than comparable females of C. tigris marmoratus This translates into larger clutches, with the mean clutch size of C. tesselatus being twice as large as that of C. tigris marmoratus The disparity in mean clutch size in conjunction with habitat constraints on C. tigris marmoratus probably explains why C. tesselatus outnumbers both C. tigris marmoratus and hybrids by a ratio of approximately 2:1 at the hybridization site. Although hybridization between C. tesselatus and C. tigris marmoratus appears to be an annual event at Arroyo del Macho, there is no evidence that a new triploid parthenogenetic species is resulting from this hybridization activity—all female hybrids examined were sterile. Nevertheless, the hybridization taking place at Arroyo del Macho is a remarkable natural experiment in progress, with either evolutionary alternative—speciation vs. destabilizing hybridization—adding to an understanding of the dynamics between parthenogenetic and bisexual species in sympatric associations.

HYBRIDIZATION AMONG WESTERN WHIPTAIL LIZARDS (CNEMIDOPHORUS TIGRIS) IN SOUTHWESTERN NEW MEXICO:POPULATION GENETICS, MORPHOLOGY, AND ECOLOGY INTHREE CONTACT ZONES

Abstract Cnemidophorus tigris punctilinealis of the Sonoran Desert and C. t. marmoratus of the Chihuahuan Desert contact each other and interbreed in the Animas Valley of southwestern New Mexico. More than 600 specimens have been examined from the contact region, and data on biochemical genetics (mitochondrial DNA haplotypes, protein electrophoresis of nuclear gene products), chromosomes, external morphology (coloration, size, scalation), reproduction, and fitness have been compared for three hybrid zones. Habitats in the contact region were mapped and photographed, and they are discussed in the context of vegetational changes during Pleistocene to Recent times, which affected the geographic distribution of these animals. Data from mitochondrial DNA, allele frequencies at four protein loci (of 36 analyzed), and body coloration demonstrate that the areas of contact have steep, concordant, and coincident step-clines in which most gene exchange occurs in hybrid zones that are 3.2–7.8 km wide. Analyses of allele frequencies, genotype frequencies, and fixation indices (including Hardy-Weinberg equilibrium, linkage equilibrium, and cytonuclear equilibrium) indicate a population structure determined primarily by random mating and an absence of selection against hybrids. Estimates of gene flow indicate that the clines resulted from neutral secondary contact initiated with the newest reconnection of the Sonoran and Chihuahuan Deserts within the present interglacial episode, from 1000 to 5000 years ago. This timeframe is consistent with paleoecological data from packrat middens. Analyses of karyotypes, morphology, reproduction, and physiology also fail to detect differences in fitness among lizards with various genotypes. Although it is possible that there are fitness differences that are too small to be detected by the sample sizes we employed, the data indicate that reproductive success, fitness, and the dynamics of populations within the hybrid zones presently are no different from those in nonhybrid populations. Earlier data, which suggested that one of the step-clines was moving, are not supported. The clines are located in fragile semiarid habitats that are subject to desertification. Consequently, we present considerable data and dated photographs of habitats, precise locations of sampling sites, and local allele frequencies, so that future investigators can monitor changes in position, width, or dynamics of these hybrid zones. In addition, the population genetics data are discussed in the context of the following: (1) absence of rare, apparently novel alleles forming in the hybrid zones; (2) genetic comparisons with additional subspecies of C. tigris (C. t. aethiops and C. t. septentrionalis); and (3) interspecific hybridization between C. tigris and other whiptail lizards of either bisexual or unisexual (parthenogenetic, clonal) species. Cnemidophorus tigris is one of the ancestors of some of the parthenogens, which are of hybrid origin, and our interest in their evolutionary history fuels our efforts to improve understanding of hybridization among whiptail lizards.

Amphibians and reptiles of Guyana, South America: illustrated keys, annotated species accounts, and a biogeographic synopsis

Abstract Guyana has a very distinctive herpetofauna. In this first ever detailed modern accounting, based on voucher specimens, we document the presence of 324 species of amphibians and reptiles in the country; 148 amphibians, 176 reptiles. Of these, we present species accounts for 317 species and color photographs of about 62% (Plates 1–40). At the rate that new species are being described and distributional records are being found for the first time, we suspect that at least 350 species will be documented in a few decades. The diverse herpetofauna includes 137 species of frogs and toads, 11 caecilians, 4 crocodylians, 4 amphisbaenians, 56 lizards, 97 snakes, and 15 turtles. Endemic species, which occur nowhere else in the world, comprise 15% of the herpetofauna. Most of the endemics are amphibians, comprising 27% of the amphibian fauna. Type localities (where the type specimens or scientific name-bearers of species were found) are located within Guyana for 24% of the herpetofauna, or 36% of the amphibians. This diverse fauna results from the geographic position of Guyana on the Guiana Shield and the isolated highlands or tepuis of the eastern part of the Pantepui Region, which are surrounded by lowland rainforest and savannas. Consequently, there is a mixture of local endemic species and widespread species characteristic of Amazonia and the Guianan Region. Although the size of this volume may mislead some people into thinking that a lot is known about the fauna of Guyana, the work has just begun. Many of the species are known from fewer than five individuals in scientific collections; for many the life history, distribution, ecology, and behavior remain poorly known; few resources in the country are devoted to developing such knowledge; and as far as we are aware, no other group of animals in the fauna of Guyana has been summarized in a volume such as this to document the biological resources. We briefly discuss aspects of biogeography, as reflected in samples collected at seven lowland sites (in rainforest, savanna, and mixed habitats below 500 m elevation) and three isolated highland sites (in montane forest and evergreen high-tepui forest above 1400 m elevation). Comparisons of these sites are preliminary because sampling of the local faunas remains incomplete. Nevertheless, it is certain that areas of about 2.5 km2 of lowland rainforest can support more than 130 species of amphibians and reptiles (perhaps actually more than 150), while many fewer species (fewer than 30 documented so far) occur in a comparable area of isolated highlands, where low temperatures, frequent cloudiness, and poor soils are relatively unfavorable for amphibians and reptiles. Furthermore, insufficient study has been done in upland sites of intermediate elevations, where lowland and highland faunas overlap significantly, although considerable work is being accomplished in Kaieteur National Park by other investigators. Comparisons of the faunas of the lowland and isolated highland sites showed that very few species occur in common in both the lowlands and isolated highlands; that those few are widespread lowland species that tolerate highland environments; that many endemic species (mostly amphibians) occur in the isolated highlands of the Pakaraima Mountains; and that each of the isolated highlands, lowland savannas, and lowland rainforests at these 10 sites have distinctive faunal elements. No two sites were identical in species composition. Much more work is needed to compare a variety of sites, and especially to incorporate upland sites of intermediate elevations in such comparisons. Five species of sea turtles utilize the limited areas of Atlantic coastal beaches to the northwest of Georgetown. All of these are listed by the International Union for the Conservation of Nature as being of global concern for long-term survival, mostly owing to human predation. The categories of Critically Endangered or Endangered are applied to four of the local sea turtles (80%). It is important to protect the few good nesting beaches for the sea turtles of Guyana. We have documented each of the species now known to comprise the herpetofauna of Guyana by citing specimens that exist in scientific collections, many of which were collected and identified by us and colleagues, including students of the University of Guyana (UG). We also re-identified many old museum specimens collected by others in the past (e.g., collections of William Beebe) and we used documented publications and collection records of colleagues, most of whom have been working more recently. We present dichotomous keys for identifying representatives of the species known to occur in Guyana, and we present brief annotated species accounts. The accounts provide the current scientific name, original name (with citation of the original description, which we personally examined in the literature), some outdated names used in the recent past, type specimens, type localities, general geographic distribution, examples of voucher specimens from Guyana, coloration in life (and often a color photograph), and comments pointing out interesting subjects for future research.

PHYSIOLOGICAL VARIATION AND ALLOMETRY IN WESTERN WHIPTAIL LIZARDS (CNEMIDOPHORUS TIGRIS) FROM A TRANSECT ACROSS A PERSISTENT HYBRID ZONE

A hybrid zone involving Cnemidophorus tigris punctilinealis (formerly gracilis) and C. tigris marmoratus in southwestern New Mexico and adjacent Arizona is narrow and characterized by abrupt and concordant change in both morphological characters and allele frequencies studied by protein electrophoresis. We compared adult C. tigris sampled from three locations that span the hybrid zone. Body mass was positively associated with both treadmill endurance at 1.0 km/h and maximal sprint running speed on a high-speed treadmill, although the largest individuals were not the fastest sprinters. Males and females differed significantly for maximal sprint ruloning speed, liver mass, and kidney mass (ANCOVA with body mass as covariate). We found no statistically significant population differences for body mass, maximal sprint running speed, standard metabolic rate at 40C, blood hematocrit levels, or heart mass. Hybrids tended to have lower treadmill endurance rlnning capacities as compared with the pure forms, but the difference was not statistically significant. Cnemidophorus tigris punctilinealis and the hybrids both had significantly heavier kidneys, relative to body mass, than did C. tigris marmoratus. Hybrid individuals also had significantly heavier livers as compared with either pure population. However, the present data cannot rule out the possibility that the observed differences in organ masses were related to reproductive status as opposed to being genetically based population differences. Thus, our results do not suggest that hybrid individuals differ from nonhybrids with respect to Darwinian fitness.

Parthenogenetic reptiles: new subjects for laboratory research.

Problems preventing establishment of laboratory colonies of parthenogenetic lizards have been solved. Now, productive colonies of these lizards, which have remarkably little genetic variation, can be readily established and used not only for research on parthenogenesis but also for many kinds of experiments for which reptile systems are desirable. Research colonies can provide valuable specimens while reducing the exploitation of natural populations.

Laboratory Hybridization Among North American Whiptail Lizards, Including Aspidoscelis Inornata Arizonae × A. tigris marmorata (Squamata: Teiidae), Ancestors of Unisexual Clones in Nature

ABSTRACT The natural origin of diploid parthenogenesis in whiptail lizards has been through interspecific hybridization. Genomes of the parthenogens indicate that they originated in one generation, as the lizards clone the F1 hybrid state. In addition, hybridization between diploid parthenogens and males of bisexual species has resulted in triploid parthenogenetic clones in nature. Consequently, the genus Aspidoscelis contains numerous gonochoristic (= bisexual) species and numerous unisexual species whose closest relatives are bisexual, and from whom they originated through instantaneous sympatric speciation and an abrupt and dramatic switch in reproductive biology.

Sexual behaviour in unisexual lizards

In captivity, females of parthenogenetic species of whiptail lizards (Cnemidophorus) occasionally mount other females and behave as if attempting to mate. This occurs under crowded conditions, and probably is not related to reproduction.

Parthenogenetic reproduction in the neotropical unisexual lizard, Gymnophthalmus underwoodi (Reptilia: Teiidae)

All adult specimens known for Gymnophthalmus underwoodi are females, and their mode of reproduction has been a mystery. In order to rule out the possibility of a bisexual mode of reproduction by means of mating with undiscovered males, hermaphroditism, or sex reversal during ontogeny, we examined hundreds of serial histological sections of complete reproductive tracts from juveniles and adults representing two generations of a lineage raised in captivity. In addition, comparative dissections were performed on other individuals and other species, and reproduction to the F5 generation was documented in laboratory colonies of G. underwoodi established from Trinidad and Surinam stocks. A lineage of three successive generations was produced entirely by individuals that were maintained in isolation from the moment of hatching.

Morphological Variation in a Unisexual Whiptail Lizard (Aspidoscelis exsanguis) and One of Its Bisexual Parental Species (Aspidoscelis inornata) (Reptilia: Squamata: Teiidae): Is the Clonal Species Less Variable?

ABSTRACT Two clonal lineages, each comprising multiple generations of unisexual A. exsanguis, were produced in the laboratory from two lizards that were collected at the same locality in the field. Based on 10 meristic and four additional characters, we assessed morphological scores and relative variation as follows: (1) between the two laboratory lineages; (2) between these lineages pooled and samples of A. exsanguis and the bisexual (gonochoristic) A. inornata from the field; and (3) between field samples of the clonal lizards and A. inornata from a nearby locality. The two lineages differed significantly in the means and variances of two univariate characters and the two most informative multivariate characters. Contrary to expectations, the pooled sample of cloned laboratory lineages of A. exsanguis were as variable as the bisexual species in all 10 univariate characters and four important multivariate characters.